Voice Recognition

As predicted yesterday, in this piece (“Typing one-handed”), I am experimenting with dictation and am currently speaking these words to my laptop. It is surprisingly easy, although I still have to go back and amend the punctuation, one-handed. Over the years, on numerous projects related to software upgrades, I have tested various versions of Dragon NaturallySpeaking, speech recognition and transcription software made by Nuance. The results were variable but generally satisfactory. There are compatibility issues with different versions of Dragon, Windows and Microsoft Office, and for the best results you need a noise-cancelling headset. The kind of headphones with built in microphone that you can pick up for 10 quid at Maplin’s don’t work as well as those supplied by Nuance.

My initial thoughts were that I would have to buy a copy of Dragon for myself to begin dictating text but I have found that for simple dictation my MacBook Pro, with its built-in microphone, works perfectly well. There is no need to install any further software. All you have to do is press the Function key (Fn) twice in any application where you can type text and start dictating. The only downside appears to be that after around 20 seconds the dictation service turns itself off, but if you press the Function key twice again it restarts. There is an option to “Use enhanced dictation”. It tells me that this “Allows offline use and continuous dictation with live feedback. Requires 422Mb download”. I’ll stick with the built-in version for now and download the enhanced version sometime later. For my current purposes this is just fine. I had no idea that it was so easy.

Clearly the world of voice recognition has moved on dramatically in the last 10 years. Millions of people are dictating instructions to Alexa or Siri or whatever other digital assistants might be out there, but I have never used any of them. I feel self-conscious enough on the phone let alone talking to a computer. I feel self-conscious now speaking these words out loud, even with nobody anywhere near me.

My experiences with different versions of Dragon over the years taught me that I have an advantage over many people when it comes to dictating text. A lot of voice recognition software responds well to my kind of voice (male, reasonably deep, London accent). If I had a high-pitched female voice with a Scottish accent the whole experience might be less satisfactory or if I put on an Australian accent the results are very different. For instance, “Streuth mate, there’s a bloke with no strides on,” a line from one of Paul Hogan’s classic Foster’s adverts, where he’s at the ballet (and I had to type those words in quotes), comes out as the following: “Street tonight there is a black with nice try it on”. This is the sort of thing that puts people off using voice recognition. I’ll stick with it, but when it comes to playing music or ordering things from Amazon I’ll do it do I always have, using my hands, not by talking to Alexa, Siri or anyone else.



One thought on “Voice Recognition

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s